
How Investors Spot Technical Risk in a Portfolio Company — Before It's Expensive
By Tom Lang on October 5, 2026 · Updated November 2, 2026
You don't need to read code to spot a portfolio company headed for an expensive technical problem. The signals are all outside the codebase, and by the time they show up in the numbers, the problem is already costly. If you're an investor or on a board, here's how to read technical risk early (including the AI risk most people aren't reading yet) without commissioning a full audit.
(I've been the CTO on the other side of these conversations, and I get placed into portfolio companies to de-risk exactly this: increasingly as AI-augmented technical due diligence across a whole portfolio. Here's what I look for first.)
Why technical risk hides
Technical risk is the last thing to appear in the metrics. Revenue looks fine, the team looks busy, the demo works: right up until a launch slips for the third time, a security review kills a deal, or the one engineer who understood everything quits. The founder often can't see it either, because it's their system and they've normalized it. So you can't wait for it to show up in the numbers; you read the leading signals.
Six signals you can read without an audit
- They can't forecast. The roadmap slips repeatedly and the explanations are vague: usually a sign no one understands the system well enough to estimate it.
- Key-person risk. If the whole thing lives in one or two heads, that's a single point of failure on your investment.
- Reliability degrading as they scale. Outages climbing with growth means they're borrowing against the product to move fast.
- Security/compliance deferred. In anything regulated, it's a landmine, and a deal-blocker at the next raise or exit when a buyer's diligence finds it.
- The tech leader can't answer strategic questions crisply, or the title doesn't match the substance (a senior engineer wearing a CTO title who can build but can't lead).
- AI risk: the new one. Is the company leaning on AI to ship fast with no one owning the standards? Unreviewed AI-generated code is tomorrow's tech debt and security exposure, accruing faster than anything we've seen before. The flip side is just as telling: a company refusing to touch AI while competitors get cheaper and faster carries a different risk. Either extreme (AI hype or AI avoidance) says there's no adult in the room on technology.
The two questions to ask
Most of this surfaces with two questions in a board meeting or diligence call:
- "What are your three biggest technical risks right now (including how you're using AI), and what are you doing about each?" A strong leader answers crisply and specifically; a weak one gets vague or defensive. The quality of the answer is the signal, as much as the content.
- "If your top engineer left tomorrow, what breaks?" If the honest answer is "everything," you've found your key-person risk.
You'll learn more from those two questions than from a stack of dashboards.
What to do when you see red: and the AI-augmented read
Seeing risk doesn't mean you need a full technical audit or a new full-time CTO tomorrow: both are slow and heavy. Usually the right move is lighter: a short, focused technical diagnostic to size the risk in writing, or placing an interim or fractional CTO into the company for a defined period to de-risk it (stabilize reliability, reduce key-person risk, get security in order), and prepare it for the next round or exit.
Here's where the economics moved in your favor. That diagnostic is now AI-augmented: the right tools read across a portfolio company's entire codebase in days, mapping the risk at a depth and speed a traditional audit can't match, while a human makes the judgment calls. That means a real technical read on several portfolio companies for what one old-school audit used to cost. For an investor, that's the leverage, one relationship, a repeatable diligence capability across the whole portfolio.
Why it matters at exit
This compounds at the worst possible moment. Technical debt, key-person risk, deferred security, and unreviewed AI slop don't just slow a company down; they become deal-blockers in diligence and get discounted straight out of the valuation. The risk you catch early and quietly fix is worth far more than the one a buyer's technical team finds for you.
What buyers now ask about AI-built code
Diligence has a new page. Buyers have started asking specifically about AI-generated code, and the questions are only answerable with records you either kept or didn't. Ownership of AI-written code isn't automatic: buyers ask for development records showing human contribution and review: source-control history, review evidence, who actually decided what. They ask for a license inventory that covers not just open-source code but datasets and model weights. They ask about the AI tools' business terms and data-retention settings, proof that proprietary code wasn't used to train someone else's model. And they ask for governance evidence: a written policy on AI use in production, a named owner for AI standards, documented human oversight.
The pattern to notice: none of this is expensive to have ready, and all of it is brutal to reconstruct in a data room. The companies that kept review discipline while they shipped fast answer in an afternoon. The ones that didn't answer with a shrug, and the buyer prices that shrug into the valuation.
For a worked example of a clean technical read, the full diligence packet on my own solo AI build (Pilot for Trades) is public: measured coverage, security posture, and an honest gap ledger (the receipts repo). It's an R&D build: never released, no customers, by design.
If a portfolio company is flashing these signals
That's exactly the kind of situation I'm brought in to read and de-risk: a single company, or AI-augmented technical due diligence across the portfolio. Book a call and we'll talk through it, or use an AI-augmented Technology Diagnostic to get one company's risk sized in writing, fast. You can also point a company at the free AI-First Readiness Scorecard for a quick self-read.
Tom Lang has led engineering organizations from founding teams to 150+ engineers across regulated, reliability-critical industries, and provides AI-augmented fractional and interim CTO leadership (including portfolio-company de-risking and technical due diligence) through Heart of the Rockies Management Consulting.

Related reading
AI-Augmented Technical Due Diligence: What Changes When the Code Was Written by AI
By Tom Lang on November 16, 2026
Technical due diligence used to mean sampling a codebase and interviewing the team. Now the target may have written half its code with AI, and the buyer may be reading it with AI too. Here's how I run AI-augmented diligence for investors: what AI changes on both sides of the table, the questions I always ask about AI-built code, and what the deliverable should look like.
Read more
Who Owns Your AI-Written Code? What Acquirers Will Ask
By Tom Lang on November 2, 2026
Buyers have started asking a question most engineering teams can't answer yet: who actually owns your AI-written code? Ownership isn't automatic, and diligence now asks for records of human review, a license inventory that includes model weights, your AI tools' data-retention terms, and governance evidence. Here's the five-part checklist to have ready before the data room opens, and why it's cheap to prepare now and brutal to reconstruct later.
Read more
The 5 Signs Your Engineering Is Quietly Burning Your Runway
By Tom Lang on October 5, 2026
Engineering rarely fails loudly. There's usually no single disaster, just a slow leak that looks like 'busy' right up until it looks like a crisis. Here are five signs, in business terms, that your engineering is costing you more than it should, and what each one actually signals. None of them are really code problems; they're judgment problems.
Read moreWant this turned into a prioritized, written plan?
That's what an AI-Augmented Technology Diagnostic is: fixed scope, fixed fee, about a week, ending in a written, prioritized plan you own. Or book a call and tell me what's going on.
